Custom Crowds in C++ for Python. A fork of the CCC software designed to be controlled from Python at the system level using Pybind11.
Designed for UNIX systems and has been tested to work on Ubuntu Linux, Ubuntu via WSL, and MacOS. Only tested using the g++ compiler.
To compile and use the program, the following must be installed:
- CMake. Set to require version 3.1 or above but easily changed by an experienced user.
- GNU Scientific Library (I used v2.4)
- Visualisation ToolKit (VTK) v6 or above is required (I have verified with v6.3 and v9.0.1)
- Python (I assume you use Python 2 but if you recquire Python 3 you can easily edit the scripts as such)
GSL and VTK can be installed on Ubuntu using the following top-level packages:
libgsl-dev
libvtk6-dev
Note that if you install VTK using apt above then you must use Python 2, as that version is configured for Python 2 only. Other versions of VTK installed via apt may work but this is untested. If you require Python 3 try installing VTK from source instead.
Output .vtp visualisation files can be read using the ParaView software.
Pybind11 is included as part of the repo.
Note that I assume you are compiling the binary for Python 2 (due to the analysis code we were using). If you are using Python 3 then you are recommended to configure VTK from source and force it to use Python 3 in cmake. You will also need to edit CMakeLists.txt on lines 15 and 18. Finally, you will need to pass CMake the location of your VTK build directory when configuring CCCpy.
The python binary is designed to be compiled via CMake. It is strongly recommended to build in a build directory. A standard build process involves calling the following from the root of the CCCpy directory:
mkdir build
cd build
cmake ..
make
This will make a compiled Python binary CCCpy(dot)so containing the System class, the main interface for running CCC simulations from Python.
Move the compiled binary to the same directory as where you intend to run your Python controlling script.
To test usage, you can try running the script at /scripts/demonstration.py which demonstrates basic usage of CCCpy - Copy the Python binary to /scripts/ and then from that directory call python demonstration.py. This script creates a subdirectory called 'Demonstration' with CCC files in "experiment format" and runs a simulation demonstrating Active Jamming behaviour (S Henkes, 2011). The .vtp files contained within the 'Demonstration/VTP/' can be opened and displayed in ParaView.
A tutorial on general usage is provided here. After importing CCCpy in Python you can call help(CCCpy) to get information on the available function calls and their respective parameters.